부산대학교 도서관

Nowadays, welding robots are widely used in the automobile manufacturing industry, but the current solder joint path planning mainly depends on the experience of technicians, and an automatic path planning method is urgently needed. The traditional algorithm can not adapt to the complex working environment of welding robot. In the intelligent algorithm, the advantage of genetic algorithm (GA) is that its overall retrieval level is high and the robustness is good. Therefore, in this research, GA is selected for path planning of welding robot. In this research, firstly, the working conditions of the welding robot are established by the grid map method, and then the adaptive level function is created by referring to the needs of its motion smoothing level and route distance, and the weighting factors of the appropriate smoothing level and route distance and the relevant parameters of the GA are found. Finally, MATLAB is used for simulation verification. Through simulation experiments, it can be concluded that the change of different parameters will have a greater impact on the path of welding robot. By adjusting the parameters, the welding robot can find the optimal path in different environments, which verifies the feasibility and effectiveness of the algorithm.